diff --git a/src/codecvt_null.cpp b/src/codecvt_null.cpp
new file mode 100644
index 0000000..624afc2
--- /dev/null
+++ b/src/codecvt_null.cpp
@@ -0,0 +1,84 @@
+/////////1/////////2/////////3/////////4/////////5/////////6/////////7/////////8
+// codecvt_null.cpp
+
+// Copyright (c) 2004 Robert Ramey, Indiana University (garcia@osl.iu.edu)
+// Andrew Lumsdaine, Indiana University (lums@osl.iu.edu). 
+// Use, modification and distribution is subject to the Boost Software
+// License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at
+// http://www.boost.org/LICENSE_1_0.txt)
+
+#define BOOST_WARCHIVE_SOURCE
+#include <boost/serialization/config.hpp>
+#include <boost/archive/codecvt_null.hpp>
+
+// codecvt implementation for passing wchar_t objects to char output
+// without any translation whatever.  Used to implement binary output
+// of wchar_t objects.
+
+namespace boost {
+namespace archive {
+
+std::codecvt_base::result
+codecvt_null<wchar_t>::do_out(
+    std::mbstate_t & /*state*/,
+    const wchar_t * first1, 
+    const wchar_t * last1,
+    const wchar_t * & next1,
+    char * first2, 
+    char * last2, 
+    char * & next2
+) const {
+    while(first1 != last1){
+        // Per std::22.2.1.5.2/2, we can store no more that
+        // last2-first2 characters. If we need to more encode
+        // next internal char type, return 'partial'.
+        if(static_cast<int>(sizeof(wchar_t)) > (last2 - first2)){
+            next1 = first1;
+            next2 = first2;
+            return std::codecvt_base::partial;
+        }
+        * reinterpret_cast<wchar_t *>(first2) = * first1++;
+        first2 += sizeof(wchar_t);
+
+    }
+    next1 = first1;
+    next2 = first2;
+    return std::codecvt_base::ok;
+}
+
+std::codecvt_base::result
+codecvt_null<wchar_t>::do_in(
+    std::mbstate_t & /*state*/,
+    const char * first1, 
+    const char * last1, 
+    const char * & next1,
+    wchar_t * first2,
+    wchar_t * last2,
+    wchar_t * & next2
+) const {
+    // Process input characters until we've run of them,
+    // or the number of remaining characters is not
+    // enough to construct another output character,
+    // or we've run out of place for output characters.
+    while(first2 != last2){
+        // Have we converted all input characters? 
+        // Return with 'ok', if so.
+        if (first1 == last1)
+             break;
+        // Do we have less input characters than needed
+        // for a single output character?        
+        if(static_cast<int>(sizeof(wchar_t)) > (last1 - first1)){
+            next1 = first1;
+            next2 = first2;
+            return std::codecvt_base::partial; 
+        }
+        *first2++ = * reinterpret_cast<const wchar_t *>(first1);
+        first1 += sizeof(wchar_t);
+    }
+    next1 = first1;
+    next2 = first2;
+    return std::codecvt_base::ok;
+}
+
+} // namespace archive
+} // namespace boost